Managerial Discretion — Personality × Power

Modules 2 through 5 taught you what CEO traits are associated with. This module teaches the condition under which any of that matters: the CEO has to have room to act.

Managerial discretion is the latitude of action available to an executive — how many plausible options exist, and how free the executive is to choose among them without being overruled by a regulator, board, owner, lender, market or culture. Where discretion is low, outcomes are mostly determined by the environment and the organization, and two very different CEOs would produce similar results. Where it is high, the CEO's dispositions are written directly into strategy, risk, capital allocation and culture.

That makes discretion the amplifier on every trait you have studied. Narcissism, overconfidence, extraversion, founder conviction: each shows up more strongly in the evidence where the CEO has more power. So do the good traits. Discretion is not a villain; it is a volume knob.

In the Effectiveness Equation — Traits × Behaviors × Organizational Context × Current Moment — this module changes the Organizational Context term, and explains why Context multiplies Traits rather than adding to them. A trait at volume zero is inaudible. A trait at full volume is the company.

You will leave with a map of where discretion comes from, an honest account of how much CEOs seem to matter at all (a contested question), and a discretion audit you can run on your own role this week.

Personality matters in proportion to power.

A trait that produces nothing in a constrained role can produce a strategy, a balance sheet and a culture in an unconstrained one. That is why the same CEO can look disciplined under a strong owner and reckless a year after that owner is gone. The question is never "what is this CEO like?" but "what is this CEO like, multiplied by how much room they have?"

FRAMEWORK The concept begins with Hambrick & Finkelstein (1987), who introduced managerial discretion as a bridge between two older camps: one holding that executives determine outcomes, the other that executives are captives of industry, structure and inertia. Both are right, in different places. Discretion derives from three sources: the task environment (industry structure, regulation, competition), the internal organization (ownership, board power, resources, culture, inertia) and the executive's own characteristics. The paper is conceptual; cite it as the origin of the idea, not as evidence for it.

RESEARCH FINDING Hambrick (2007), updating upper echelons theory (Hambrick & Mason, 1984), formalized discretion as a moderator: executive characteristics predict organizational outcomes only where executives have latitude. He added a second moderator — executive job demands — arguing that heavy demands push executives toward heuristics and personal disposition; pressure makes CEOs more like themselves. This too is a conceptual review.

RESEARCH FINDING Wangrow, Schepker & Barker (2015) reviewed the empirical work and concluded the theory remains incompletely validated: most studies operationalize discretion through industry and task-environment factors, far fewer measure organizational or individual sources, measurement is inconsistent, and it is unclear whether constraints substitute or compound. INTERPRETATION The strongest evidence you have concerns environmental discretion — the industry you are in — and the weakest concerns the personal sources executives most like to believe in.

RESEARCH FINDING The most direct test of Personality × Power is Li & Tang (2010). In a survey of 2,790 CEOs of Chinese manufacturing firms, survey-measured CEO hubris was positively related to firm risk taking, and the relationship was markedly stronger when the CEO had more discretion. Amplifying conditions included munificent but complex markets, low organizational inertia, greater intangible resources, CEO–board-chair duality and CEOs who were not politically appointed. The design is cross-sectional, survey-based and single-country, with possible common-source concerns; the exact hubris measure is not described here because it was not independently confirmed. But the pattern is what the theory predicts: hubris plus latitude produces risk.

RESEARCH FINDING Chatterjee & Hambrick (2007) studied 111 CEOs in computer hardware and software (1992–2004) using an unobtrusive narcissism index built from proxies — photo prominence in the annual report, prominence in press releases, first-person-singular pronouns, and pay relative to the second-highest-paid executive. Narcissistic CEOs made more and larger acquisitions and produced more extreme, fluctuating performance, with no better or worse average performance. INTERPRETATION A single fast-moving industry is the kind of setting where discretion logic says traits should be most visible; the variance-not-mean result is what a trait does when it has room to operate.

RESEARCH FINDING Buyl, Boone & Wade (2019) bring governance in explicitly. Among 92 U.S. commercial-bank CEOs (2006–2014), archival narcissism markers predicted riskier policies before the 2008 shock, especially when pay was option-heavy — and strong board monitoring dampened the relationship. Banks led by more narcissistic CEOs also recovered more slowly. Small sample, one industry, one crisis, observational — but the shape matters: the trait's expression depends on the incentive and monitoring around it.

RESEARCH FINDING Malhotra, Reus, Zhu & Roelofsen (2018), analyzing unscripted remarks by 2,381 S&P 1500 CEOs, found that more extraverted CEOs make more and larger acquisitions, with the effect stronger in less competitive industries and where managerial entrenchment is higher. Malmendier & Tate (2009) show a related pattern: the post-award underperformance of "superstar" CEOs is strongest in firms with weak governance. Miller (1991), in a cross-sectional study whose sample size is not confirmed, found long-tenured CEOs less likely to have organizations aligned with their environments, especially under concentrated ownership. Bertrand & Schoar (2003), following executives across firms, found persistent manager "styles" in corporate policy, with better-performing styles earning more in better-governed firms.

RESEARCH FINDING— CONTESTED. How much do CEOs matter at all? Quigley & Hambrick (2015) partitioned performance variance for 1,015 U.S. firms in 30 industries (18,467 firm-years, 2,732 CEOs, 1950–2009) and found the share attributable to individual CEOs rose across three periods: by sequential ANOVA, the CEO effect on ROA went from 7.8% (1950–69) to 10.8% (1970–89) to 15.7% (1990–2009); by multilevel modeling, from 13.8% to 17.8% to 22.9%. Fitza (2014) argued that such estimates are inflated by chance: with tenures averaging about four years, random performance alone produces a sizable apparent CEO effect, because luck does not have time to average out. A press summary of his 2017 follow-up states that over 70% of the measured CEO effect could be attributable to chance. Quigley & Graffin (2017) replied that Fitza's assumptions overstate the random component and reaffirmed an effect "significant and much larger than chance"; Fitza (2017) rejoined that under more realistic assumptions the effect remains close to chance regardless of estimator. The debate is unresolved. Any single percentage should carry the Fitza caveat.

Where the evidence is weak

Discretion is more often assumed than measured. The trait × discretion interactions rely on proxies for the trait (photos, pronouns, pay gaps, option holding, speech) and proxies for discretion (duality, competition, entrenchment). Designs are associations, not experiments; CEOs with strong traits may select into high-discretion settings rather than being amplified by them. The CEO-effect literature measures a statistical residual associated with CEO identity, not any specific action. Treat this module as a well-supported direction with poorly measured magnitudes.

Where discretion comes from

FRAMEWORK Following Hambrick & Finkelstein's three sources, it helps to break discretion into eleven practical constraints. Each is a place where someone or something other than the CEO decides, or narrows the menu.

From the environment: regulation (a bank CEO cannot decide their own capital ratio), industry structure (a commodity producer cannot set price), capital intensity (a steel mill takes six years and $2B; a software product takes six months), competition (a rival's move forces a response) and geography (labor law, currency, political risk).

From the organization: corporate governance (charter, bylaws, the audit and risk machinery), board power (an active, independent board narrows discretion; a captive one widens it), ownership concentration (a controlling owner or sponsor can veto anything; dispersed owners mostly cannot), debt covenants (a hard wall), investor expectations (guidance, the narrative the CEO has sold, the activist on the register) and organizational culture (what the CEO's orders actually mean by the time they reach the floor).

These are not all the same kind of constraint. Some are legal walls (covenants, regulation), some political (board, owner), some economic (capital intensity, competition), some invisible until you push on them (culture). Wangrow et al. (2015) could not tell whether constraints substitute or compound. HYPOTHESIS In practice they seem to compound at the low end — a regulated, covenant-bound, owner-controlled company gives its CEO very little — and substitute at the high end, where removing the last strong constraint (a founder-chair, a lead lender, a large shareholder) opens the whole field at once.

Why traits need power to be visible

Imagine two CEOs with identical, strongly aggressive dispositions. One runs a regional electricity utility under a rate commission, with a sixty-year asset base and two former regulators on the board. The other runs a founder-controlled, cash-rich software company with a three-person board and no debt. The first CEO's aggression shows up, if at all, as a faster construction schedule and a combative tone in rate hearings. The second CEO's aggression shows up as the company's strategy.

Upper echelons theory (Hambrick & Mason, 1984) says firms reflect their top managers. Discretion says: only to the extent that the managers are allowed to do the reflecting.

FRAMEWORK— Personality × Power. The product form is deliberate. Add the terms and a strong personality always matters somewhat. Multiply them and you get the world the evidence describes: near-zero discretion makes even a strong trait nearly invisible; high discretion makes even a moderate trait consequential. The uncomfortable corollary: the CEOs whose personalities you can see clearly in biographies are, disproportionately, the ones who had unusual power. Their stories are about Personality × Power, not personality.

Why this matters most for the dangerous traits

INTERPRETATION Module 4 established that narcissism and overconfidence raise the variance of outcomes more reliably than the mean (Chatterjee & Hambrick, 2007; Cragun et al., 2020). Discretion converts that variance from a statistical abstraction into a company-sized bet. A narcissistic CEO in a low-discretion role makes bolder speeches. A narcissistic CEO with duality, a friendly board and a war chest makes the acquisitions. Li & Tang's (2010) hubris finding, Buyl et al.'s (2019) monitoring finding and Malhotra et al.'s (2018) entrenchment finding say the same thing from different data: the risk-taking of dispositionally bold CEOs is conditional on how little stands in their way.

Founder control deserves its own line. A founder with a large equity stake, a hand-picked board and a company built in their image has, structurally, the highest discretion any CEO gets. Module 9 covered why that can be a genuine advantage. This module adds the multiplier: whatever a founder is — visionary or grandiose, patient or impulsive — the company will be that, at full volume. The founder's traits are moderated by nothing except the founder's own maturity, which is why Level 3 of the Maturity Model is not optional for founders. There is no one else to do it.

Dominant CEOs in professional settings are the same case with a different origin story. A CEO who has accumulated the chair role, a long tenure, a board they recruited and a string of wins has manufactured founder-level discretion without founding anything. Miller (1991) found long tenure associated with worse organization–environment fit, particularly under concentrated ownership.

Job demands: pressure turns the volume up further

Hambrick (2007) added a moderator most practitioners miss: under heavy job demands, executives rely more on heuristics and personal disposition. INTERPRETATION So the moments when discretion is widest (a crisis in which the board defers, an owner's death, a windfall) are often the moments when the CEO is most stressed and therefore most themselves. If you want to know what a CEO is really like, watch the first ninety days after a constraint disappears.

How much do CEOs matter? Less than the press implies; more than nothing; and it depends

RESEARCH FINDING— CONTESTED. Even the most generous estimate in the CEO-effect debate — 22.9% of performance variance in the most recent period by multilevel modeling (Quigley & Hambrick, 2015) — leaves more than three-quarters to industry, firm, year and noise, and Fitza (2014, 2017) argues that much of what remains is luck wearing a CEO's name tag. INTERPRETATION Two implications: humility about attribution (when a company performs well, neither outsiders nor the CEO can tell how much was the CEO), and leverage — if CEOs matter mainly through discretion, then how much discretion to grant a CEO is one of the highest-leverage decisions a board makes.

The discretion audit

FRAMEWORK Score your own role. For each of the eleven sources, assign 0 (binds me hard), 1 (shapes what I do but I have real room), or 2 (barely constrains me). Be honest about formal versus effective constraint: a board that approves everything is a 2, not a 0.

1. Regulation. Could a regulator veto my three most important decisions this year?
2. Industry structure. Do I set price, or does the market?
3. Capital intensity. Can I change what the company does within eighteen months?
4. Competition. Does a rival's move force my hand?
5. Geography. Do local law, labor, currency or politics narrow my choices?
6. Governance machinery. Do the audit, risk and compensation processes actually stop things?
7. Board power. When did the board last say no to me? Be specific.
8. Ownership concentration. Can one person remove me in a phone call?
9. Debt covenants. Would my preferred plan breach one?
10. Investor expectations. Have I sold a narrative I would be punished for abandoning?
11. Culture. Do my decisions execute as intended, or does the organization quietly reinterpret them?

Total the score. Zero to seven: a constrained role; your judgment matters, your personality mostly does not. Eight to fourteen: moderate discretion; your dials matter, and so does coalition-building. Fifteen to twenty-two: founder-level discretion; your traits are the company's traits, and whatever you overuse, the company will overuse. Above fourteen, the most important question is the follow-up: which constraints have I removed myself, and why?

Fictional composite.

Corrigan Building Supply was a $420M-revenue distributor of lumber, roofing and windows, with 1,900 employees and 38 branches across four states. Founded in 1971 and wholly owned by Walt Corrigan, who chaired the board at 79, it had been run day to day for nine years by Dale Whitcomb, a non-family CEO promoted from COO.

Under Walt, the constraints were simple and absolute. Any acquisition over $5M needed Walt's signature. Net debt could not exceed 1.5 times EBITDA — a rule Walt adopted after nearly losing the company in a 1990 housing slump. The board was Walt, Dale, Walt's lawyer and the company's long-time banker.

Dale was, by temperament, an aggressive, fast, optimistic operator who had wanted to roll up smaller distributors for years. Under Walt he had done two small deals, both under the cap, both integrated well. The company grew about 6% a year at steady 4.1% operating margins, and the bank and the industry regarded Dale as a disciplined, conservative CEO. On the discretion audit, his role scored a five.

Walt died in March. Ownership passed to a trust for three adult children, none of whom worked in the business. The board became Dale, the lawyer and two heirs. Within a month the acquisition cap was gone — no one had thought to keep it — and the leverage rule became advisory. The bank, dealing now with a CEO it trusted and heirs it barely knew, extended a $120M acquisition facility. Dale's discretion audit went from five to about eighteen in six weeks. Nothing about Dale changed.

Over the next twenty months Dale acquired three regional distributors for a combined $140M, taking net debt to 3.4 times EBITDA. The logic was defensible: purchasing scale, delivery density. But each deal was done fast and priced to win. Integration ran late; two acquired management teams left. The heirs, receiving quarterly decks and a confident narrative, asked few questions; the lawyer asked none. Then regional housing starts fell 11%, the largest acquired business missed plan by 30%, and Corrigan breached its coverage covenant. The bank imposed a chief restructuring officer; the heirs appointed two independent directors and reinstated a leverage limit. Dale stayed, chastened, and stabilized the company over three years.

Dale did not become reckless; he had always been aggressive. Walt's constraints made his aggression invisible and, in two small deals, useful. When the constraints vanished, the same dial setting produced a three-deal, $140M, 3.4x-levered strategy. Personality × Power: the personality was constant; the power tripled.

It could have gone the other way. Had housing held and the integrations succeeded, Dale would have been the CEO who "finally had room to run." That is the variance point. Discretion did not make Dale wrong; it made Dale consequential. The board's failure was not choosing Dale — it was removing every constraint at once without asking what Dale was like at full volume.

Put four archetypes in Dale's chair the week after Walt's funeral, with the same suddenly-open field.

The Operator CEO would notice the open field and be uneasy about it. Operators sit toward operational discipline and caution; their instinct when constraints vanish is to replace them. This CEO would likely propose to the heirs a written capital policy, a leverage ceiling and an independent director before any acquisition — partly from conviction, partly because discipline is easier to keep when it is someone else's rule. Cost: slower consolidation; a competitor may buy the best targets. Gain: risk bounded by design rather than luck. The danger is the opposite of Dale's — using self-imposed constraint as a reason never to do the deal that scale actually required.

The Visionary CEO would treat Walt's death as the moment the company could finally become what it should have been. Optimism and aggression far left, urgency high. This CEO would move faster than Dale and be more persuasive with the heirs, because visionaries are good at narrative. Gain: if the thesis is right, this is the CEO who builds a $1B platform. Cost: the widest variance of the four; with no board that can say no, the outcome is almost entirely a function of whether the thesis is right — the bet Chatterjee & Hambrick's (2007) findings warn about. This archetype most needs the constraint it least wants.

The PE-Backed CEO has lived inside sponsor governance — monthly board packs, a 100-day plan, an operating partner, real covenants. Dropped into a constraint-free family company, this CEO would feel the absence of structure as a risk and would likely rebuild it: an independent chair, covenants, a deal committee. Gain: professional governance quickly. Cost: the CEO's own risk appetite — often high, since sponsors select for it — is now bounded only by rules the CEO wrote.

The Family Business CEO (imagine a Corrigan heir who did work in the business) carries an internal constraint the others lack: identity. This CEO would likely keep Walt's leverage rule out of loyalty as much as prudence, and consolidate slowly if at all. Gain: the lowest chance of a covenant breach. Cost: discretion high on paper and low in practice — constrained by legacy and the fear of being the one who lost it — which can leave the company under-invested exactly when scale matters. Miller's (1991) concentrated-ownership finding is the warning.

Discretion does not have a good or bad setting. It has a right size for the CEO's temperament and maturity, and the board's job is to size it.

FRAMEWORK— the Overuse Ladder under high discretion. Confidence → arrogance, risk tolerance → recklessness, decisiveness → impulsiveness, dominance → intimidation, vision → fantasy. Discretion is how a CEO climbs the ladder without noticing, because every rung requires that no one stops you. In a constrained role, arrogance is corrected by the board, recklessness by the covenant, impulsiveness by the process. Remove the constraints and the ladder has no rails.

The destructive pattern is usually not a single reckless act. It is the slow, rational-seeming removal of constraints by a CEO who has been right several times in a row. Each removal is defensible: the deal cap is slowing us down; the chair role will let me move faster; the covenants are legacy terms from a weaker period. Hayward, Rindova & Pollock (2004) theorize that celebrity-induced hubris leads CEOs to persist in the actions that produced their fame; Chatterjee & Hambrick (2011) find narcissistic CEOs respond to praise more than to results. INTERPRETATION Add discretion and you get the standard trajectory: success → praise → confidence → constraint removal → larger bets → higher variance. The CEO experiences this as earned freedom. The organization experiences it as a widening cone of outcomes.

Early warning signs

For the CEO:

• You cannot remember the last time the board said no, and you regard approval as a formality to be managed.
• Proposals arrive pre-shaped to what you are known to want.
• You have argued for combining chair and CEO, or for not replacing the director who pushes back, and your stated reasons are about speed.

For the board:

• Risk limits, capital policies and deal caps have been loosened one at a time, each with a good reason, and no one has looked at the cumulative picture.
• Board information is narrative-heavy and more optimistic in tone than the numbers.
• The CEO's largest recent decisions coincide with praise, awards or press attention.
• Nobody could name the constraint that would stop the CEO from doing the one thing that could sink the company.

The remedy is rarely to remove the CEO. It is to re-size the discretion: an independent chair or empowered lead director, restored limits, a deal committee, a risk appetite statement with teeth, and — hardest — a culture in which the CEO's proposals meet informed disagreement. Buyl et al. (2019) found strong board monitoring dampened the narcissism–risk link in banks. Monitoring is not an insult to the CEO. It is the rail on the ladder.

1. Run the discretion audit on your current role. What is your total? Which two constraints did you have to think hardest about?
2. Which constraints have you personally removed or weakened in the last three years? For each, write the reason you gave at the time and the reason a skeptic would give.
3. Name the last decision a board, owner or lender stopped you from making. If you cannot, what does that say about your discretion — and the quality of your information?
4. Recall a time your behavior looked different in a constrained role versus a freer one. Which version was closer to your temperament? Which produced better results?
5. If your discretion doubled tomorrow, which Overuse Ladder rung would the company climb first? What would the first symptom be?
6. How much of your company's performance in the last three years do you honestly attribute to your decisions versus industry, timing and luck? Ask a director the same question about you and compare.
7. What constraint would you want if you knew you were about to become far more powerful? Why have you not built it already?

• Managerial discretion — the latitude to act — is the condition under which any CEO trait matters. Personality × Power, not personality, predicts outcomes.
• The best-supported sources of discretion are environmental (industry, regulation, competition); organizational and individual sources are less studied. Use the discretion audit to estimate your own.
• The variance-raising traits — narcissism, overconfidence, hubris, extraversion — are amplified by discretion in study after study and dampened by monitoring, competition and constraint. Founder control and CEO–chair duality are the highest-discretion settings.
• The size of the CEO effect is contested; even generous estimates leave most performance variance to industry, firm and chance. Attribute your own success with humility.
• The common path to failure is the gradual, well-argued removal of constraints by a CEO who has been right several times. The remedy is to re-size discretion, not to pretend it does not matter.